Gurusha Juneja

Resume | SOP | Scholar | Github | Email

I am a first year PhD student an University of California, Santa Barbara, to be advised by Prof. William Wang. Previously I was a Research Fellow at Microsoft Research India, working with Dr. Amit Sharma. I obtained my Bachelor's degree in EE from IIT-Delhi. Here, I worked with Prof. Tanmoy Chakraborty.

Looking for summer internships in areas related to agents, reasoning and planning, multi-modal AI, or AI for science.

My research interests lie in understanding and developing AI architectures that can perform scientific reasoning and long horizon planning reliably. A related topic that hugely interests me in the development of world models and verifiers. Two directions I am currently exploring are - 1. understanding Multi-Agent systems and their efficient use; 2. Text Diffusion for iterative solution refinement. I am also interested in exploring reasoning in latent space.

Looking for strong graduate/undergraduate students to collaborate. Please reach out if you are interested in agents/diffusion related topics.

News

April, 2024: New paper on solving complex reasoning using a society of models is out !
Feburary, 2024: Will present our paper at AAAI 2024, Canada
December, 2023: Our work Prompt-Propose-Verify gets accepted as a Oral presentation in AAAI workshop AI for Digital Human
December, 2023: Presented my poster at EMNLP 2023, Singapore
November, 2023: Awarded Microsoft Research Travel Grant and Diversity and Inclusion Award to attend EMNLP 2023
October, 2023: Our work UniPrompt gets accepted as a spotlight paper in NeurIPS workshop R0-FoMo
October, 2023: Our work DaSLaM gets accepted in the Conference Proceedings of EMNLP 2023 Main
August, 2023: I will be reviewing for EMNLP'23!

Publications

LM2: A Simple Society of Language Models Solves Complex Reasoning

Gurusha Juneja, Subhabrata Dutta, Tanmoy Chakraborty

TACL Main 2024 | Paper

Small Language Models Fine-tuned to Coordinate Larger Language Models improve Complex Reasoning

Gurusha Juneja, Subhabrata Dutta, Soumen Chakrabarti, Sunny Manchanda, Tanmoy Chakraborty

EMNLP Main 2023 | Paper | Code

A Universal Prompt Generator for Large Language Models

Gurusha Juneja*, Amit Sharma

NeurIPS 2023 Workshop | Paper

Prompt-Propose-Verify: A Reliable Synthetic Data Generation Framework using Foundational Models

Gurusha Juneja, Sukrit Kumar

AAAI 2023 Workshop | Paper | Code